Robotics
PyTorch
Cosmos
xperience10m_task_baseline_suite
embodied-ai
multimodal
xperience-10m
baseline
evaluation
qwen3-omni
Instructions to use cy0307/ropedia-xperience-10m-task-baselines with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- Cosmos
How to use cy0307/ropedia-xperience-10m-task-baselines with Cosmos:
# No code snippets available yet for this library. # To use this model, check the repository files and the library's documentation. # Want to help? PRs adding snippets are welcome at: # https://github.com/huggingface/huggingface.js
- Notebooks
- Google Colab
- Kaggle
| #!/usr/bin/env python3 | |
| """Build deterministic brand assets from the generated logo mark.""" | |
| from __future__ import annotations | |
| import hashlib | |
| import json | |
| from datetime import datetime, timezone | |
| from pathlib import Path | |
| from PIL import Image, ImageDraw, ImageFilter, ImageFont | |
| ROOT = Path(__file__).resolve().parents[1] | |
| BRAND_DIR = ROOT / "docs/assets/brand" | |
| SOURCE_MARK = BRAND_DIR / "xperience10m-logo-mark.png" | |
| OUTPUT_JSON = ROOT / "docs/data/brand_assets.json" | |
| OUTPUTS = { | |
| "mark_512": BRAND_DIR / "xperience10m-logo-mark-512.png", | |
| "mark_192": BRAND_DIR / "xperience10m-logo-mark-192.png", | |
| "favicon_64": BRAND_DIR / "xperience10m-logo-favicon-64.png", | |
| "favicon_32": BRAND_DIR / "xperience10m-logo-favicon-32.png", | |
| "apple_touch": BRAND_DIR / "xperience10m-logo-apple-touch.png", | |
| "social_card": BRAND_DIR / "xperience10m-logo-social-card.png", | |
| "root_favicon": ROOT / "docs/favicon.png", | |
| "root_apple_touch": ROOT / "docs/apple-touch-icon.png", | |
| } | |
| INK = (244, 248, 239) | |
| MUTED = (190, 202, 184) | |
| GREEN = (167, 240, 120) | |
| CYAN = (122, 229, 195) | |
| BG = (2, 5, 2) | |
| PANEL = (5, 14, 8) | |
| LINE = (43, 92, 41) | |
| def resample(): | |
| return getattr(Image, "Resampling", Image).LANCZOS | |
| def load_font(size: int, *, bold: bool = False) -> ImageFont.FreeTypeFont | ImageFont.ImageFont: | |
| candidates = [ | |
| "/System/Library/Fonts/Supplemental/Arial Bold.ttf" if bold else "/System/Library/Fonts/Supplemental/Arial.ttf", | |
| "/System/Library/Fonts/Supplemental/Helvetica Bold.ttf" if bold else "/System/Library/Fonts/Supplemental/Helvetica.ttf", | |
| "/Library/Fonts/Arial Unicode.ttf", | |
| ] | |
| for candidate in candidates: | |
| path = Path(candidate) | |
| if path.exists(): | |
| return ImageFont.truetype(str(path), size=size) | |
| return ImageFont.load_default() | |
| def sha256(path: Path) -> str: | |
| digest = hashlib.sha256() | |
| with path.open("rb") as handle: | |
| for chunk in iter(lambda: handle.read(1024 * 1024), b""): | |
| digest.update(chunk) | |
| return digest.hexdigest() | |
| def image_record(name: str, path: Path, role: str) -> dict: | |
| with Image.open(path) as image: | |
| return { | |
| "name": name, | |
| "path": path.relative_to(ROOT).as_posix(), | |
| "role": role, | |
| "exists": path.exists(), | |
| "bytes": path.stat().st_size, | |
| "sha256": sha256(path), | |
| "format": image.format, | |
| "width": int(image.width), | |
| "height": int(image.height), | |
| "mode": image.mode, | |
| } | |
| def alpha_crop(image: Image.Image, padding_ratio: float = 0.08) -> Image.Image: | |
| rgba = image.convert("RGBA") | |
| alpha = rgba.getchannel("A") | |
| bbox = alpha.getbbox() | |
| if bbox is None: | |
| raise ValueError(f"No visible pixels in {SOURCE_MARK}") | |
| left, top, right, bottom = bbox | |
| width = right - left | |
| height = bottom - top | |
| pad = int(max(width, height) * padding_ratio) | |
| left = max(0, left - pad) | |
| top = max(0, top - pad) | |
| right = min(rgba.width, right + pad) | |
| bottom = min(rgba.height, bottom + pad) | |
| return rgba.crop((left, top, right, bottom)) | |
| def fit_on_canvas(image: Image.Image, size: int, *, scale: float = 0.88) -> Image.Image: | |
| canvas = Image.new("RGBA", (size, size), (0, 0, 0, 0)) | |
| cropped = alpha_crop(image) | |
| max_side = int(size * scale) | |
| cropped.thumbnail((max_side, max_side), resample()) | |
| x = (size - cropped.width) // 2 | |
| y = (size - cropped.height) // 2 | |
| canvas.alpha_composite(cropped, (x, y)) | |
| return canvas | |
| def make_dark_tile(mark: Image.Image, size: int) -> Image.Image: | |
| tile = Image.new("RGBA", (size, size), (*BG, 255)) | |
| glow = Image.new("RGBA", (size, size), (0, 0, 0, 0)) | |
| glow_draw = ImageDraw.Draw(glow) | |
| glow_inset = max(2, size // 18) | |
| glow_draw.rounded_rectangle( | |
| (glow_inset, glow_inset, size - glow_inset - 1, size - glow_inset - 1), | |
| radius=max(5, size // 7), | |
| fill=(*GREEN, 82), | |
| ) | |
| glow_draw.ellipse( | |
| (size * 0.22, size * 0.22, size * 0.78, size * 0.78), | |
| fill=(*CYAN, 42), | |
| ) | |
| glow = glow.filter(ImageFilter.GaussianBlur(max(2, size // 10))) | |
| tile = Image.alpha_composite(tile, glow) | |
| draw = ImageDraw.Draw(tile) | |
| border_width = max(2, size // 30) | |
| draw.rounded_rectangle( | |
| (1, 1, size - 2, size - 2), | |
| radius=max(4, size // 8), | |
| fill=(2, 15, 7, 250), | |
| outline=(*GREEN, 245), | |
| width=border_width, | |
| ) | |
| inner_inset = max(4, border_width + size // 16) | |
| draw.rounded_rectangle( | |
| (inner_inset, inner_inset, size - inner_inset - 1, size - inner_inset - 1), | |
| radius=max(2, size // 10), | |
| outline=(*CYAN, 92), | |
| width=max(1, size // 80), | |
| ) | |
| fitted = fit_on_canvas(mark, size, scale=0.87) | |
| tile.alpha_composite(fitted) | |
| return tile | |
| def draw_grid(draw: ImageDraw.ImageDraw, width: int, height: int) -> None: | |
| step = 34 | |
| for x in range(0, width, step): | |
| for y in range(0, height, step): | |
| if (x // step + y // step) % 3 == 0: | |
| draw.ellipse((x, y, x + 2, y + 2), fill=(35, 72, 34)) | |
| def make_social_card(mark: Image.Image) -> Image.Image: | |
| width, height = 1200, 630 | |
| card = Image.new("RGB", (width, height), BG) | |
| draw = ImageDraw.Draw(card) | |
| draw_grid(draw, width, height) | |
| glow = Image.new("RGBA", (width, height), (0, 0, 0, 0)) | |
| glow_draw = ImageDraw.Draw(glow) | |
| glow_draw.ellipse((38, 66, 548, 576), fill=(38, 108, 42, 78)) | |
| glow_draw.ellipse((120, 148, 466, 494), fill=(122, 229, 195, 34)) | |
| glow = glow.filter(ImageFilter.GaussianBlur(34)) | |
| card = Image.alpha_composite(card.convert("RGBA"), glow) | |
| panel_glow = Image.new("RGBA", (470, 470), (0, 0, 0, 0)) | |
| panel_glow_draw = ImageDraw.Draw(panel_glow) | |
| panel_glow_draw.rounded_rectangle((24, 24, 446, 446), radius=38, fill=(*GREEN, 56)) | |
| panel_glow_draw.rounded_rectangle((54, 54, 416, 416), radius=32, fill=(*CYAN, 28)) | |
| panel_glow = panel_glow.filter(ImageFilter.GaussianBlur(22)) | |
| card.alpha_composite(panel_glow, (61, 80)) | |
| panel = Image.new("RGBA", (420, 420), (0, 0, 0, 0)) | |
| panel_draw = ImageDraw.Draw(panel) | |
| panel_draw.rounded_rectangle( | |
| (0, 0, 419, 419), | |
| radius=34, | |
| fill=(3, 15, 8, 238), | |
| outline=(*GREEN, 238), | |
| width=3, | |
| ) | |
| panel_draw.rounded_rectangle( | |
| (16, 16, 403, 403), | |
| radius=26, | |
| outline=(*CYAN, 92), | |
| width=1, | |
| ) | |
| mark_fit = fit_on_canvas(mark, 390, scale=0.9) | |
| panel.alpha_composite(mark_fit, (15, 15)) | |
| card.alpha_composite(panel, (86, 105)) | |
| title_font = load_font(64, bold=True) | |
| subtitle_font = load_font(36, bold=True) | |
| body_font = load_font(25) | |
| small_font = load_font(22) | |
| mono_font = load_font(20, bold=True) | |
| x = 570 | |
| draw = ImageDraw.Draw(card) | |
| draw.text((x, 145), "Ropedia", font=title_font, fill=INK) | |
| draw.text((x, 218), "Xperience-10M", font=title_font, fill=GREEN) | |
| draw.text((x, 308), "Task Suite", font=subtitle_font, fill=CYAN) | |
| draw.text( | |
| (x, 370), | |
| "Multimodal embodied-AI task baselines", | |
| font=body_font, | |
| fill=MUTED, | |
| ) | |
| badge_y = 448 | |
| badges = ["video", "audio", "depth", "pose", "mocap", "IMU", "language"] | |
| cursor = x | |
| row = 0 | |
| max_x = width - 86 | |
| for badge in badges: | |
| label_width = int(draw.textlength(badge, font=mono_font)) | |
| if cursor + label_width + 28 > max_x: | |
| row += 1 | |
| cursor = x | |
| y = badge_y + row * 46 | |
| box = (cursor, y, cursor + label_width + 28, y + 36) | |
| draw.rounded_rectangle(box, radius=9, fill=(7, 22, 12), outline=LINE, width=1) | |
| draw.text((cursor + 14, y + 8), badge, font=mono_font, fill=INK) | |
| cursor += label_width + 40 | |
| draw.line((x, 555, width - 86, 555), fill=(78, 151, 72), width=1) | |
| draw.text((x, 573), "single-sample evidence now | multi-episode fine-tuning next", font=small_font, fill=(155, 170, 149)) | |
| return card.convert("RGB") | |
| def main() -> int: | |
| if not SOURCE_MARK.exists(): | |
| raise FileNotFoundError(f"Missing source logo mark: {SOURCE_MARK}") | |
| BRAND_DIR.mkdir(parents=True, exist_ok=True) | |
| mark = Image.open(SOURCE_MARK).convert("RGBA") | |
| fit_on_canvas(mark, 512).save(OUTPUTS["mark_512"]) | |
| fit_on_canvas(mark, 192).save(OUTPUTS["mark_192"]) | |
| make_dark_tile(mark, 64).save(OUTPUTS["favicon_64"]) | |
| make_dark_tile(mark, 32).save(OUTPUTS["favicon_32"]) | |
| make_dark_tile(mark, 180).save(OUTPUTS["apple_touch"]) | |
| make_dark_tile(mark, 64).save(OUTPUTS["root_favicon"]) | |
| make_dark_tile(mark, 180).save(OUTPUTS["root_apple_touch"]) | |
| make_social_card(mark).save(OUTPUTS["social_card"], optimize=True, quality=92) | |
| manifest = { | |
| "title": "Ropedia Xperience-10M Brand Assets", | |
| "status": "pass", | |
| "generated_at_utc": datetime.now(timezone.utc).isoformat(timespec="seconds"), | |
| "source": { | |
| "path": SOURCE_MARK.relative_to(ROOT).as_posix(), | |
| "kind": "custom generated logo mark with chroma-key background removed locally", | |
| "prompt_summary": "X-shaped multimodal camera mark with near-black, lime, cyan, trajectory, and point-cloud styling.", | |
| }, | |
| "assets": [ | |
| image_record("logo_mark", SOURCE_MARK, "Transparent source logo mark."), | |
| image_record("logo_mark_512", OUTPUTS["mark_512"], "512px transparent logo mark."), | |
| image_record("logo_mark_192", OUTPUTS["mark_192"], "192px transparent logo mark for app manifest use."), | |
| image_record("favicon_64", OUTPUTS["favicon_64"], "64px dark-tile favicon and navigation logo."), | |
| image_record("favicon_32", OUTPUTS["favicon_32"], "32px dark-tile favicon fallback."), | |
| image_record("apple_touch", OUTPUTS["apple_touch"], "180px apple-touch icon."), | |
| image_record("social_card", OUTPUTS["social_card"], "1200x630 Open Graph, Twitter, README, and HF-card logo card."), | |
| image_record("root_favicon", OUTPUTS["root_favicon"], "Root website favicon PNG."), | |
| image_record("root_apple_touch", OUTPUTS["root_apple_touch"], "Root website apple-touch icon."), | |
| ], | |
| "boundary": "Brand assets are generated presentation artifacts. They do not contain raw Xperience-10M video, HDF5, RRD data, or model weights.", | |
| } | |
| OUTPUT_JSON.parent.mkdir(parents=True, exist_ok=True) | |
| OUTPUT_JSON.write_text(json.dumps(manifest, indent=2) + "\n", encoding="utf-8") | |
| for name, path in OUTPUTS.items(): | |
| print(f"{name}: {path} ({path.stat().st_size} bytes)") | |
| print(f"manifest: {OUTPUT_JSON} ({OUTPUT_JSON.stat().st_size} bytes)") | |
| return 0 | |
| if __name__ == "__main__": | |
| raise SystemExit(main()) | |